knowledge factuality benchmark
May 2025 revision; 2158 text-only questions; no tools; 1 run; pass@1; GPT-5.6 Luna medium judge
Updated Sep 22, 2026
Higher hle ranks better on this benchmark.
Rank | Model | hle | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAN | hle59.13% | Percentile100.00% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank02 | ModelAN | hle55.47% | Percentile99.50% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank03 | ModelAN | hle54.87% | Percentile99.00% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank04 | ModelOP | hle54.68% | Percentile98.49% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank05 | ModelOP | hle49.49% | Percentile97.99% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank06 | ModelME | hle48.70% | Percentile97.49% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank07 | ModelAN | hle48.66% | Percentile96.98% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank08 | ModelGO | hle47.87% | Percentile96.48% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank09 | ModelGO | hle47.82% | Percentile95.98% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank10 | ModelGO | hle47.03% | Percentile95.48% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank11 | ModelMA | hle46.90% | Percentile94.97% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank12 | ModelME | hle46.20% | Percentile94.47% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank13 | ModelOP | hle45.78% | Percentile93.97% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank14 | ModelME | hle45.46% | Percentile93.47% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank15 | ModelXA | hle44.07% | Percentile92.96% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank16 | ModelOP | hle43.74% | Percentile92.46% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank17 | ModelAC | hle43.10% | Percentile91.96% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank18 | ModelOP | hle42.91% | Percentile91.46% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank19 | ModelGO | hle42.68% | Percentile90.95% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank20 | ModelXA | hle42.68% | Percentile90.45% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank21 | ModelOP | hle42.49% | Percentile89.95% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank22 | ModelAN | hle42.31% | Percentile89.45% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank23 | ModelZA | hle42.26% | Percentile88.94% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank24 | ModelAN | hle41.29% | Percentile88.44% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank25 | ModelZA | hle41.15% | Percentile87.94% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank26 | ModelDE | hle41.01% | Percentile87.44% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank27 | ModelGO | hle40.82% | Percentile86.93% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank28 | ModelME | hle40.69% | Percentile86.43% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank29 | ModelAC | hle40.50% | Percentile85.93% | Participants200 | EvidenceB | EvaluatedN/A |
| Rank30 | ModelZA | hle39.85% | Percentile85.43% | Participants200 | EvidenceB | EvaluatedN/A |
The leading models and scores on this benchmark.
A closer view of the leading scores on this benchmark.
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What AA HLE Text No Tools measures and how its scores work.
May 2025 revision; 2158 text-only questions; no tools; 1 run; pass@1; GPT-5.6 Luna medium judge
Scores are shown in ratio. This benchmark is verified and has an evidence level of B.
Artificial Analysis. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
Common questions about AA HLE Text No Tools.
Claude Fable 5.1 is currently ranked first with 59.13%.
The current leaders are Claude Fable 5.1 (59.13%), Claude Fable 5 (55.47%), and Claude Opus 5 (54.87%).
Step-3.5-Flash has the lowest matched official input price at $0.10 input / $0.30 output per 1M tokens.
The fastest matched records are Mercury 2 (1,009.00 tok/s via Inception), GPT OSS 120B (500.00 tok/s via Groq), and GPT-5 mini (348.85 tok/s via OpenAI).
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.
May 2025 revision; 2158 text-only questions; no tools; 1 run; pass@1; GPT-5.6 Luna medium judge
Yes. Higher values rank better for this benchmark.
100 model results are currently shown.
Yes. This benchmark can contribute to the current LLMBoard capability score.
Ranking basisThis aa hle text no tools AI model leaderboard uses descending hle in the benchmark original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.
Selection summary
Claude Fable 5.1 currently leads AA HLE Text No Tools with 59.13%. It is the top model on this specific benchmark, while the best LLM for the broader task should also be checked against other benchmarks, price, and runtime.
Use this leaderboard with the supporting benchmark results and coverage details above. A leaderboard position summarizes the selected ranking signal; it does not replace workload-specific testing.